Founding Agent Engineer

Optimized builds AI agents that run inside the most critical supply chains in the world, from the factory floor to the procurement teams at Fortune 500 manufacturers and government organizations. An agent is only as good as the decisions it makes, and we need a founding engineer to own that intelligence.

As a Founding Agent Engineer, you'll define how our agents reason, plan, and decide: the prompting, retrieval, tool-use strategy, and fine-tuning that turn raw model capability into good supply chain decisions. You'll own decision quality end-to-end, from research bench to production.

Compensation range for this role is $150,000 - $220,000 + equity.

• Own how agents reason:
  You'll design the reasoning, planning, and tool-use strategies that let our agents scout suppliers, model costs, and assess risk, and you'll be accountable for the quality of those decisions.
• Go deep on the models:
  You'll own prompting, retrieval-augmented generation, and fine-tuning, pushing the limits of what the underlying models can reliably do.
• Make decision quality measurable:
  You'll build the eval frameworks and feedback loops that tell us whether an agent is actually making better decisions, and raise the bar on measuring before optimizing.
• Embed with customers:
  You'll spend time inside real supply chain operations, turning messy, high-stakes problems into tractable reasoning tasks for our agents.
• Set the intelligence bar:
  You'll work directly with the founders to define how our agents think, and set the example for what world-class applied AI looks like here.

• Have 4+ years building production AI/ML systems, including owning model behavior end-to-end from research to deployment
• Have hands-on experience with LLMs in production (prompting, retrieval-augmented generation, fine-tuning, and evaluation)
• Are fluent in Python and modern ML tooling (PyTorch, Hugging Face, or similar)
• Are rigorous about evaluation, and you measure decision quality before you optimize
• Are willing to work hard, move fast, and grow quickly in a rapidly changing environment
• Are excited about applying AI to real-world industrial problems, not just benchmarks